Promote “Best Practices” in Medicine:

“Pay for Performance” is a common objective for many leaders in healthcare public policy. The goal of this effort is to encourage physicians who practice the highest quality medicine and achieve the best possible outcomes. Those who fall short do not receive full payment by Medicare. However, this policy has many unintended consequence.

Example #1:

Data reporting the number of deaths at a given hospital are becoming available to the public. On first glance, a hospital with the lower death rate appears to “perform” better than a competing hospital with a higher death rate. However, this is not necessarily the case.

If a hospital with relatively poor care transfers their sickest patients to another facility, their death rate will appear very low. Conversely, an excellent hospital that accepts the sickest patients from outside facilities may in fact see the number of deaths go up. Though the numbers are supposed to be adjusted for severity of illness, the experience of practicing physicians suggests these numbers are highly unreliable.

Example #2:

“Pay for performance” can actually adversely impact patient care and increase healthcare costs. When asked, families sometimes request that a family member only receive conservative care. However, “pay for performance” can discourage physicians from having difficult end-of-life discussions.

For example children sometimes request that their 95-year-old, bed-bound father with severe debilitation and advanced Alzheimer’s dementia receive only IV fluids and antibiotics. They often request that if their father’s heart or breathing stops that he not have CPR and not be placed on life support. His comfort is their primary concern.

A hospital or nursing home physician who routinely makes time for these difficult end-of-life discussions will have more deaths than a physician who always uses every resource available without making time to talk to patients and families. “Pay for performance” may actually discourage these discussions and force physicians to either “do everything” and return the patient to the nursing home or to “do nothing” and place the patient on hospice so the death does not count against the hospitals mortality rate.

Example #3:

Data reporting the number of post-operative complications (such as infections) of a given surgeon are becoming available to the public. Again, on first glance this appears to be a useful indicator of a surgeon’s skill. One would think patients seen by the best surgeons would have the fewest complications.

However, this system discourages surgeons from operating on patients most likely to experience complications. Higher risk patients such as those who are morbidly obese, critically ill, or have advanced diabetes or heart disease may find surgeons less likely to perform a needed procedure.

Example #4:

Driven by “pay for performance” concerns, hospitals now review charts looking to see if patients with a diagnosis of congestive heart failure have number called the “ejection fraction” documented on the chart. This number represents how efficiently the heart pumps blood and is most often obtained by with a study called an echocardiogram.

While useful in many situations, an echocardiogram is not always needed. For example, a patient with a history of congestive heart failure may have shortness of breath from pneumonia, not a new problem with his or her heart. However, because this is one of the ways Medicare measures “performance,” physicians sometimes order unnecessary echocardiograms. Ordering this study is easier than fighting a system to explain why it does not need to be done.

Example #5:

Driven by “pay for performance,” hospitals now receive something called a “Press-Ganey” score which measures patient satisfaction. Medicare uses this score to reduce hospital reimbursement for hospitals that are not in the upper (approximately) tenth percentile.

Because a growing number of physicians are going out of business, more physicians are becoming hospital employees. Because the hospitals financing depends on how patients rate their physician, some hospitals are linking physicians salaries to their Press-Ganey score.

Driven by this distortion of the normal patient / physician relationship, physicians are adjusting their behaviors. There are now reports of emergency room physicians increasing the amount of narcotics prescribed in an attempt to keep drug-seeking patients from returning a low score.

Conclusion:

Though it appears to make sense on the surface, using “pay for performance” to threaten physicians and hospitals does not always result in better healthcare. The real goal of “pay for performance” is to encourage physicians and hospitals to practice the best possible medicine. However, it often only serves to make a given statistic look better.

A Solution:

Medical societies from nearly every specialty are developing what are know as “best practices” for illnesses within that specialty. In short, “best practices” represent the most efficient and effective way known to treat a given condition, as determined by physicians.

Linking “best practices” to medical litigation reform would incentivize physicians and hospitals to practice better medicine. This would encourage better performance without the unintended consequences of “pay for performance.” Statistical analysis is a powerful tool for observing trends and finding correlations. However, as seen above, statics often lose their value when used to influence behavior.

How can tort reform encourage the use of “best practices”? If a physician accurately diagnoses a patient and follows the recommended “best practice” for that condition, he or she would be protected from frivolous litigation. (This protection would not apply for gross negligence not related to the given “best practice.”)

Bt itself, “pay for performance” ultimately encourages physicians to change behavior in a way that makes a given statistic look better. Linking “best practices” to tort reform take the enormous amount of energy surrounding healthcare litigation and redirects it toward implementing the best practices of medicine.